AI models
Transformer-XL + RMT
Training compute
6.8×10¹⁸ FLOP
Parameters
247M
Published
Jul 14, 2022
Transformer-XL + RMT is an AI model developed by Moscow Institute of Physics and Technology and AIRI Artificial Intelligence Research Institute (Russia), first published in July 2022. It works in the language domain, on tasks such as language modeling.
Training it took an estimated 6.8×10¹⁸ FLOP of compute (estimation method: operation counting,hardware). The model has 247,000,000 parameters. It was trained on roughly 103M datapoints. Training ran on 2 NVIDIA A100 SXM4 80 GB for about 30 hours.
Access: Unreleased. Its weights are not openly released. The reference paper has 183 citations. Epoch AI rates the confidence of this record as speculative.
Full record
- Organization
- Moscow Institute of Physics and Technology, AIRI Artificial Intelligence Research Institute
- Country of organization
- Russia
- Domain
- Language
- Task
- Language modeling
- Training compute
- 6.8×10¹⁸ FLOP
- Compute estimation method
- Operation counting, Hardware
- Parameters
- 247,000,000
- Dataset size
- 103M
- Training hardware
- NVIDIA A100 SXM4 80 GB
- Chips used
- 2
- Training time
- 30 h
- Training power draw
- 1.6 kW
- Model accessibility
- Unreleased
- Open weights
- No
- Citations
- 183
- Epoch confidence
- Speculative
SourceEpoch AI, 'AI Models'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/ai-models. Licensed under CC BY 4.0.
← All ai models